Many optical image applications require very fast large aperture image shutters. Mechanical shutters, though available with almost any size aperture, are generally too slow for these applications. Gated image intensifiers are fast enough, having rise times on the order of nanoseconds but they require generally undesirably large driving voltages that are in the range of 160 to 1000 volts. Liquid crystal shutters operate at low voltages and though they are significantly faster than mechanical shutters they are still much slower than gated image intensifiers.
Low voltage high speed optical modulators or shutters that operate at frequencies on the order of 10.sup.9 Hertz have been described in: "High Speed Optical Modulation with GaAs/GaAlAs Quantum Wells in a P-I-N Diode Structure", T. H. Woods et al, Appl. Phys. Lett. 44(1), p16 (1984); "Band-Edge Electro- /absorption in Quantum Well Structures: The Quantum--Confined Stark Effect", D. A. B. Miller et al, Phys. Rev. Lettr. 26, p2173 (1984); "Electric Field Dependence of Optical Absorption Near Band Gap of Quantum Well Structures", D. A. B. Miller et al, Phys. Rev. B, vol 32 #2, p1043, (1985); and U.S. Pat. No. 4,525,687 to Chemla et al., all of which are incorporated herein by reference. These modulators are based on the fact that the absorptive part (the imaginary part) of the index of refraction of light in a semiconductor can be made dependent upon an electric field applied to the material of the semiconductor. The shutters are made from very thin layers of semiconductor material that are sandwiched together to form a series of quantum wells. The quantum well structure of the shutters amplifies the effect of the electric field on the absorptive part of the index of refraction of the material and thereby provides relatively large changes in absorption of light for relatively small operating voltages. While these semiconductor shutters are very fast and operate at low voltages they achieve their speed because their active volumes and apertures are very small.
They are used primarily for optical communications applications where their small apertures are acceptable.
The technology of semiconductor optical shutters is based on the way photons interact with semiconductor materials. As a beam of light travels through a semiconductor, the photons in the beam interact with the material of the semiconductor and are absorbed from the beam causing the light beam to be attenuated. If the rate of attenuation of photons in the semiconductor is graphed as a function of the energy of the photons, the attenuation rate generally shows a very steep rise, called the absorption edge of the semiconductor, at a well-defined photon energy or wavelength. The energy, hereafter the "absorption edge energy", at which the absorption edge occurs is generally very close to the band gap of the semiconductor. Photons having energy below the absorption edge energy (and therefore a wavelength above the wavelength corresponding to the absorption edge energy, hereafter the "absorption edge wavelength"), interact very weakly (if at all) with the semiconductor material and are only very slightly attenuated by the semiconductor material. Photons with energy above the absorption edge energy (i.e. wavelength below the absorption edge wavelength) interact strongly with the material and are rapidly attenuated per unit path length in the material.
Typically, the absorption length for photons of energy just below the absorption edge energy is on the order of 10 cm.sup.-1 and for photons of energy just above the absorption edge energy, .about.10.sup.4 cm.sup.-1. Generally, the absorption edge energy can be shifted to lower energies by applying an electric field to the semiconductor and then can be shifted back to higher energies by removal or reduction of the applied electric field. As a result, for photons of energy sufficiently close to the absorption edge energy the absorption edge energy can be shifted to just below or just above the energy of the photons by an electric field applied to the semiconductor. This causes the photons to be either very strongly or very weakly absorbed by the semiconductor. In this way, the amount of light transmitted by the semiconductor material at a wavelength, hereafter an operational wavelength, close to the absorption edge wavelength of the semiconductor can be controlled by an electric field in the semiconductor. The semiconductor acts like an electrically operated optical shutter for light having a wavelength equal to an operational wavelength of the semiconductor material.
The absorption edge of a semiconductor occurs at an energy slightly less than the band gap energy of the semiconductor. Instead of lifting an electron from the valence band into the conduction band, a photon can excite an electron almost to the conduction band and leave it loosely bound to the hole from which it was lifted in a short lived Hydrogen like resonance. The bound resonance of electron and hole is called an exciton. The binding energy of the exciton is typically low, on the order of a few mev. At room temperature an exciton quickly picks up energy from phonons and ionizes into an uncoupled electron and hole, with the electron entering the conduction band. If the energy at which the absorption edge occurs is "Ex", and the band gap energy and exciton binding energies are "Eg" and "Eb" respectively then Ex=Eg-Eb.
When an electric field is applied to a bulk semiconductor, both Eg and Eb are broadened and reduced. The electric field broadens and reduces Eg by coupling energy to valence electrons over distances on the order of the cell length of the semiconductor (typically 0.5 nm), in what is known as the Franz-Keldysh effect. The electric field broadens and reduces exciton binding energies, Eb, by coupling energy to excitons over distances on the order of the size of their diameters, which is typically about 30 nm. The effect of an electric field on an exciton is a Stark effect, whereby the electric field slightly polarizes the exciton and increases the average separation of the hole and electron in the exciton, thereby reducing the binding energy of the exciton.
Neither the Franz-Keldysh effect nor the Stark effect is effective for shifting the absorption edge energy Ex, in a bulk semiconductor. The Franz-Keldysh effect is too small. The Stark effect, while significantly larger, rapidly ionizes the exciton which causes the energy of the absorption edge to be insensitive to an applied electric field.
The effect of electric fields on absorption edge energy increases significantly when the semiconductor is formed from thin layers of narrow band gap semiconductor material alternating with layers of wide gap semiconductor material, where the thickness of the narrow gap layers is significantly less than the diameter of excitons. In such a layered structure, the narrow gap layers act as quantum wells having physical widths equal to the thickness of the narrow gap layers. Electrons and holes formed in the semiconductor material of the wells are trapped in these quantum wells at discrete energy levels. If the wide gap layers are sufficiently thick, the wave functions of electrons and holes in a quantum well are strictly confined within the quantum wells and do not tunnel through to adjacent quantum wells. The structure then behaves like a series of uncoupled quantum wells conventionally called multiple quantum wells or MQW's.
As a result of the confinement of electrons and holes in the quantum wells, the distance between the electron and hole in an exciton is restricted in a direction perpendicular to the planes of the layers to a distance less than the width of the quantum wells i.e. to the thickness of the narrow gap layers. This causes the average distance between an electron and hole in an exciton trapped in a quantum well to be smaller than the average distance between an electron and hole in an exciton in a bulk semiconductor formed from the same semiconductor material from which the quantum wells are formed. Excitons in an MQW made of a given semiconductor material therefore have a higher binding energy and are more stable than excitons in a bulk semiconductor made from the same given material. Furthermore, an electric field applied perpendicular to the MQW layers will not polarize and distort excitons to such an extent (it cannot separate an electron and hole in an exciton by more than the width of the quantum well) that the excitons rapidly ionize. As a result, in a semiconductor MQW structure, unlike in a bulk semiconductor, an electric field can be used to effectively control the absorption edge energy, provided the electric field is perpendicular to the planes of the quantum wells. An MQW structure can therefore be used as an optical shutter.
If the discrete energies of the first confined states of electrons and holes in the quantum wells are respectively "Ee" and "Eh", then the absorption edge energy for the MQW's, is Ex=Eg+Ee+Eh-Eb. An electric field applied to the MQW structure that significantly affects any of the energies Eg, Eh or Eb can be used to shift Ex and effectively control the absorption edge energy and wavelength. For an MQW in which the narrow gap layers are formed from GaAs and the wide gap layers are formed from Ga.sub.(1-x) Al.sub.x As with x .about.0.36, and the thickness of the layers is approximately 10 nm, Ee, Eh, and Eb are on the order of 50 mev, 5 mev, and 10 mev respectively and Ex is .about.1.45 ev. An electric field of 1-5.times.10.sup.4 V/cm applied perpendicular to the layers of this MQW structure couples energy in the range from 10-50 mev to the electron and hole wave functions ([1-5.times.10.sup.4 V/cm.times.10.sup.-6 cm] x the charge of the electron) and can therefore be used to control the energy of the absorption edge.
US Pat. No. 4,525,687 describes an MQW optical shutter comprising 50 GaAs quantum wells each 9.5 nm thick. The wells are formed in the intrinsic part of a pin diode. The diode is operated back biased so that substantially all of the potential difference applied to the diode appears across the quantum wells. Light passes through the diode perpendicular to the planes of the quantum wells. At a wavelength of 857 nm (1.446 ev) the shutter transmits about 75% of the light when open (no voltage applied to the diode) and about 35-40% when the shutter is closed by applying a potential of 8V to the diode. The shutter therefore has an on/off transmission ratio of about 2. The useful optical aperture of the shutter described was confined to an area of about 100 microns in diameter and the active volume of the shutter to about 100 cubic microns. The shutter operates with a rise time on the order of 2 nanoseconds.
Prior art optical shutters based on MQW technology have been very small aperture, small volume devices used mainly in optical communications applications. While the shutters operate at relatively high frequencies, their apertures are too small for many applications. In particular they are much too small for shuttering full images. For example, PCT Patent Publication WO97/01111 entitled "Improved Optical Ranging Camera" published Jan. 9, 1997 whose disclosure is incorporated herein by reference, describes range finding cameras that acquire depth or distance information simultaneously over entire extended images. Such cameras require high speed, large aperture shutters that have high on/off transmission ratios. It would be desirable to have large aperture shutters for shuttering or modulating images that operate at the frequencies of the small volume MQW shutters used in optical communications and that have high on/off transmission ratios.